When fitting a linear trend to time series data we have to use a time index variable as a predictor (i.e. explanatory variable), because (select all that apply): it is very easy to relate back the time index values to the actual time units at which the data has been observed. we use a time index in regression analysis of any kind of data. in many cases due to the complex time scale of recording the observations, nothing else makes sense to use as a variable representing time. we want the intercept of the resulting linear trend model to have a meaningful interpretation. .